Direct reductive amination of ketones: Structure and activity of S-selective imine reductases from streptomyces

Article abstract of DOI:10.1002/cctc.201402218

The importance and structural diversity of chiral amines is well-demonstrated by the myriad nonenzymatic methods for their chemical production. In nature, the production of amines is performed by transamination rather than by reduction of an imine precursor derived from the corresponding ketone. Imine reductases, however, show great potential in the reduction of cyclic imines that are stable towards hydrolysis in aqueous reaction media. Here, we report the catalytic activity of two S-selective imine reductases towards 3,4-dihydroisoquinolines and 3,4-dihydro-β-carbolines and their activity in the direct reductive amination of ketone substrates. The crystal structures of the enzyme from Streptomyces sp. GF3546 in complex with the cofactor NADPH and from Streptomyces aurantiacus in native form have been solved and refined to a resolution of 1.9 A.

Full text of DOI:10.1002/cctc.201402218

CHEMCATCHEM
COMMUNICATIONS
DOI: 10.1002/cctc.201402218
Direct Reductive Amination of Ketones: Structure and
Activity of S-Selective Imine Reductases from Streptomyces
Tobias Huber,[a] Lisa Schneider,[b] Andreas Prꢀg,[a] Stefan Gerhardt,[b] Oliver Einsle,[b] and
Michael Mꢁller*[a]
The importance and structural diversity of chiral amines is
well-demonstrated by the myriad nonenzymatic methods for
their chemical production. In nature, the production of amines
is performed by transamination rather than by reduction of an
imine precursor derived from the corresponding ketone. Imine
reductases, however, show great potential in the reduction of
cyclic imines that are stable towards hydrolysis in aqueous re-
action media. Here, we report the catalytic activity of two S-se-
lective imine reductases towards 3,4-dihydroisoquinolines and
3,4-dihydro-b-carbolines and their activity in the direct reduc-
tive amination of ketone substrates. The crystal structures of
the enzyme from Streptomyces sp. GF3546 in complex with the
cofactor NADPH and from Streptomyces aurantiacus in native
form have been solved and refined to a resolution of 1.9 ꢂ.
vored equilibrium, demanding a great excess of the amine
donor or a complex multienzyme network, are general draw-
backs of the transaminase approach, making an alternative
biocatalytic solution desirable.
Amino acid dehydrogenases display reductive amination ac-
tivity towards a-keto acids[9] and, for the conversion of ketones
in the presence of ammonia, they have been engineered to
act as amine dehydrogenases; hence, the corresponding imine
is reduced to its amine by consumption of NAD(P)H.[10] Al-
though the excellent stereopreference derived from the amino
acid dehydrogenase scaffold could be retained, so far this
strategy has suffered from a poor substrate scope. In the same
manner in which advancement of a lactate dehydrogenase is
unlikely to lead to a superior alcohol dehydrogenase, we
postulate that the most promising approach for the biocatalyt-
ic production of chiral amines is the use of enzymes that act
on imines as their physiological substrates (Scheme 1a). Re-
cently, an example of a closely related activity was found in
some Streptomyces strains with the reduction of 2-methyl-1-
pyrroline (1) to (R)- or (S)-2-methylpyrrolidine (Scheme 1b).[11]
Some knowledge concerning the participating imine reduc-
tases has already been gained: the R-selective isoenzymes
from Streptomyces sp. GF3587[12] and Streptomyces kanamyceti-
cus[13] have been heterologously produced and purified, and
the crystal structure of the latter enzyme has been solved.[13]
The NADPH-dependent active protein is a homodimer of two
32 kDa subunits with a C-terminal helical domain and an
N-terminal Rossmann fold. Although the stereoselectivity of
the reduction of cyclic imine 1 is excellent, catalytic activity
towards further substrates seems to be poor.
In addition to the elucidation of the fundamental mechanisms
of metabolic processes in organisms, another aim of biocata-
lytic research is to transfer the processes evolved in nature to
organic synthesis. This approach is particularly valuable as
chemical conversions in living organisms are often similar to
the demands of the chemical industry.[1] Consequently, a variety
of enzyme-catalyzed synthetic routes have found application
in chemical synthesis.[2] Explicitly for the production of chiral
compounds, biocatalytic methods serve as an appropriate
option due to their common inherent stereoselectivity.
Due to the structural diversity of chiral amines, existing pro-
cesses for their production have been optimized and new
methods developed over the past decades.[3] In particular, bio-
catalytic strategies have been the subject of recent research.
Next to chiral resolution using hydrolases,[4] monoamine oxy-
genases,[5] and transaminases,[6] asymmetric synthesis has been
performed with transaminases.[7] The latter has reached an
impressive level of applicability.[8] Product inhibition and a disfa-
The S-selective enzyme from Streptomyces sp. GF3546[14] is
also a homodimer of two 31 kDa subunits bearing a Rossmann
fold on the N-terminal domain. The amino acid sequence
homology of the two stereocomplementary Streptomyces
imine reductases (sp. GF3546 and sp. GF3587) is 37%, with the
highest consensus in the N-terminal domain. Moreover, the S-
selective imine reductase from Streptomyces sp. GF3546 shows
a high-sequence homology to putative proteins from numer-
ous bacteria.[15]
[a] T. Huber, Dr. A. Prꢀg, Prof. Dr. M. Mꢁller
Institute of Pharmaceutical Sciences
Albert-Ludwigs-Universitꢀt Freiburg
Albertstrasse 25, 79104 Freiburg (Germany)
E-mail: michael.mueller@pharmazie.uni-freiburg.de
Based on the potential presence of an enzyme class with
the putatively native function of reducing C=N bonds, we
were interested in the characterization of such enzymes with
the aim of applying them to the production of chiral amines
by reductive amination. Herein, we describe the direct reduc-
tive amination of ketones by utilizing S-selective imine reduc-
tases from Streptomyces as catalysts for the production of
chiral amines.
[b] L. Schneider, Dr. S. Gerhardt, Prof. Dr. O. Einsle
Institute of Biochemistry
Albert-Ludwigs-Universitꢀt Freiburg
Albertstrasse 21, 79104 Freiburg (Germany)
and
BIOSS Centre for Biological Signalling Studies
Hebelstrasse 25, 79104 Freiburg (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402218.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
conformation of the enzyme
(root mean squared displace-
ments 0.38 ꢂ); in contrast, the
cofactor NADPH is only observed
in one of the monomers of the
Streptomyces sp. GF3546 protein
(Figure 1a), in which its binding
leads to a closure of the two do-
mains with respect to the unli-
ganded structure from S. auran-
tiacus, a situation that has not
been noted previously (Fig-
ure 1b). Binding of the dinucleo-
tide is mediated by the N-termi-
nal domain that adopts a canoni-
cal Rossmann fold, whereas the
C-terminal, a-helical domain is
responsible for the formation of
the highly stable dimeric form of
the imine reductase. The helices
of the two monomers are highly
Scheme 1. a) Direct biocatalytic reductive amination versus reduction of ketones. b) S-Selective imine reductase
catalyzed reduction of cyclic imines 1 and 2.
We investigated the two homologous imine reductases from
Streptomyces sp. GF3546 and Streptomyces aurantiacus
JA 4570, which show an amino acid sequence identity of 57%.
For optimal production and isolation of imine reductase from
Streptomyces sp. GF3546, a synthetic gene with codon usage
optimized for Escherichia coli was designed on the basis of
published data.[16] The gene coding for the imine reductase
from S. aurantiacus was amplified from genomic DNA[17] by
using polymerase chain reaction techniques. Both genes were
cloned into a pET22b(+) vector construct by utilizing the re-
striction sites BamHI and NdeI and overexpressed in E. coli
BL21(DE3) to give the C-terminal hexahistidine-tagged fusion
protein. The proteins were purified by affinity chromatography
(on nickel(II) nitrilotriacetic acid) and the imine reductase activi-
ty was determined by the reduction of the five-membered
cyclic imine 1. The highest specific activities of 0.035 UmgÀ1
(Streptomyces sp. GF3546) and 0.075 UmgÀ1 (S. aurantiacus)
were measured at 10 mm substrate concentration. Surprisingly,
the catalytic activity towards the six-membered cyclic imine 6-
methyl-2,3,4,5-tetrahydropyridine (2) was more than one order
of magnitude higher for both isoenzymes. A higher specific ac-
intercalated to maximize the interaction surface. Upon binding
of NADPH to one monomer of the Streptomyces sp. GF3546
enzyme, the two domains undergo a slight relative rotation
that narrows the cleft between them and forms an obvious
tivity of the imine reductase from S. aurantiacus (1.25 UmgÀ1
compared to that from Streptomyces sp. GF3546 (0.37 UmgÀ1
)
)
was also confirmed for this substrate. Maximal enzymatic activ-
ity was reached at a substrate concentration of 15 mm.
To investigate a possible structural basis for the observed
subtle differences in activity, both imine reductases have been
crystallized and their three-dimensional structures solved by
molecular replacement (Figure 1). In accordance with their
high homology in the primary sequence, the two structures
are very similar, with root mean squared displacements for all
atoms of 1.08 and 1.14 ꢂ, respectively, for the two chains of
the dimeric enzyme. They also show homology to enzyme
Q1EQE0 from S. kanamyceticus (PDB ID: 3ZGY, 3ZHB).[13] In the
S. kanamyceticus protein structures the symmetric binding of
NADPH to both subunits only leads to minor changes in the
Figure 1. Three-dimensional structures of imine reductases: a) Representa-
tion of the dimer of imine reductase from Streptomyces sp. GF3546 with
a solvent-accessible surface. Only one monomer binds a molecule of NADPH
in a cleft between the Rossmann-like nucleotide-binding domain and the
helical dimerization domain. b) Stereo representation of a superposition of
imine reductase from Streptomyces sp. GF3546 (c) and S. aurantiacus
(c). In the upper domain, nucleotide binding leads to a noticeable confor-
mational change.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
Table 1. Biocatalytic reduction of 3,4-dihydroisoquinolines 5a,b and 3,4-
dihydro-b-carbolines 7a,b by imine reductases.
Streptomyces sp. GF3546
S. aurantiacus
Conversion[a] [%]
Conversion[a] [%]
ee[b]
ee[b]
5a
5b
7a
7b
>95
67
>95
>95
98
15
98
99
>95
44
91
99
34
99
98
50
[a] Conversion after incubation for 20 h. [b] Determined by HPLC analysis.
Figure 2. Binding of NADPH: a) The nucleotide-binding cleft in Streptomyces
sp. GF3546 imine reductase in stereo representation. The structure suggests
a sequential mechanism in which binding of NADPH must precede substrate
binding for catalysis to occur. b) Stereo representation of the binding mode
of NADPH and the hydrogen-bonding interactions of the cofactor with the
surrounding protein. Residues N33, R34, T35, and K38 interact with the
2’-phosphate group at the ADP moiety of NADPH to ensure discrimination
from NADH. The Fo-Fc difference electron density map is calculated without
contributions from the bound ligand and contoured at the 3s level.
catalytic properties of the imine reductase from S. aurantiacus.
The small structural differences of the two proteins, however,
are reflected in some differences in their catalytic activities
(Table 1).
Whereas conversions and ee values for the methyl deriva-
tives 5a and 7a were essentially identical, the ethyl-substitut-
ed derivatives 5b and 7b showed distinct deviations. The ee
for the reduction of 7b to (S)-8b remained on an equally high
level. For S. aurantiacus imine reductase, however, conversion
dropped to 50%, whereas the Streptomyces sp. GF3546 imine
reductase-catalyzed process occurred quantitatively. The differ-
ence in catalytic properties became even more notable for the
formation of 6b: although the S stereopreference was re-
tained, the ee dramatically decreased (15 and 34% ee, Table 1).
After confirmation of the enzymatic activity for the stereose-
lective reduction of a range of cyclic imine substrates, the next
topic addressed was the ability of the enzymes to withdraw an
imine substrate from the equilibrium with a ketone in the pres-
ence of a primary amine or ammonia (Scheme 1a). Thus, a reac-
tion system was applied in which a ketone was incubated in
methylammonium buffer together with imine reductase,
NADP+, and a cofactor regeneration system [d-glucose with
glucose dehydrogenase (GDH)]. For this reaction setup, it was
crucial that the imine reductase itself showed no concurrent
ketoreductase activity towards the employed ketone substrate,
which was verified by blank experiments.
binding site for the secondary substrate that is to be reduced
(Figure 2a).
NADPH is coordinated to the protein through a network of
hydrogen bonds that strongly resembles the binding mode
observed for S. kanamyceticus Q1EQE0.[13] It ensures that the Si
face of the nicotinamide moiety faces the active site cleft (Fig-
ure 2b). The structure as such strongly suggests a sequential
binding mode for the nicotinamide dinucleotide and the imine
substrate, and the combination of the two structures described
here highlights a conformational change that occurs upon
NADPH binding. Although the catalytic cleft remains rather
spacious even after cofactor binding, the structural data indi-
cate that the NADPH-bound structure is the one to investigate
for possible site-directed alterations that affect the range of
substrates applicable for biocatalytic conversions.
An important aspect for the application of a biocatalyst is
the determination of the substrate scope. For imine reductas-
es, this examination is tedious as most imines are sensitive to
hydrolysis in the aqueous reaction medium. Nevertheless, 3,4-
dihydroisoquinolines (5) and 3,4-dihydro-b-carbolines (7) are
classes of compounds that are stable in water, bear a C=N
bond, and are available by known synthetic methods.[18]
In agreement with results published during our investiga-
tions, showing the catalytic activity of the Streptomyces sp.
GF3546 imine reductase towards a broader range of cyclic
imines,[19] we have confirmed that imine reductases accept this
class of imine compounds in general. This is validated by the
We could not detect any amine product on application of
the imine reductases as crude cell-free extracts. This probably
owed to the concurrent ketoreductase side-activity of soluble
proteins from E. coli, yielding up to 10% of the alcohol byprod-
uct. For the purified imine reductase from Streptomyces sp.
GF3546, the conversion of 4-phenyl-2-butanone (9) into 2-
(methylamino)-4-phenylbutane (11, Scheme 2) was (8.8Æ1.6)%
if 2.5 mg protein was used on a reaction scale of 20 mmol of
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
sion rates at different pH values and neither of the applied
imine reductases showing a clear superiority.
Although the conversions of ketone into the corresponding
amine are low, these examples represent the first direct reduc-
tive aminations of ketones by native enzymes. So far, this cata-
lytic activity has not been observed in nature, as amines are
derived through different metabolic pathways. Considering
that these two isoenzymes only represent a small selection of
an enzyme family widespread in bacterial strains and have
individual catalytic behavior towards different substrates, this
work will be a valuable starting point for further detailed inves-
tigations of direct biocatalytic reductive amination. Considering
that protein engineering is a successful tool for optimizing
enzymes with respect to reaction rates, stereoselectivity, and
substrate tolerance, such a biocatalytic process holds great
potential to extend the available methods for the production
of chiral amines.
Scheme 2. Direct biocatalytic reductive amination of ketone 9.
substrate. Notably, the catalytic activity was strongly pH
dependent and reached its maximum at pH 9.5. Such a high
pH for the maximal conversion could owe to the pH depend-
ency of imine 10 formation.
Experimental Section
3,4-Dihydroisoquinolines 5 and 3,4-dihydro-b-carbolines 7 were
produced by the Bischler–Napieralski cyclization of the correspond-
ing amides derived from phenethylamine and tryptamine, respec-
tively. The amides were heated in toluene in the presence of
phosphoryl trichloride (T=115 8C) or neat polyphosphoric acid (T=
2008C). Hydrolysis (addition of water), basification to pH 11 (addi-
tion of NaOH solution), and extraction with dichloromethane gave
the crude products, which were purified by column chromatogra-
phy. Biocatalytic reductions of 5 and 7 were performed at a
substrate concentration of 20 mm in 4-(2-hydroxyethyl)-1-piper-
azineethanesulfonic acid buffer (100 mm, pH 7.0) containing d-glu-
cose (100 mm, 5 equiv.), NADP+ (5 mol%), GDH (5.8–20 U), and
imine reductase (4 UmmolÀ1, relative to 1) in a total volume of
10 mL. After the mixture was shaken for 20 h at RT, the reaction
was terminated by the addition of 10wt.% NaOH solution (1 mL)
and the aqueous mixture was extracted with dichloromethane.
The turnover number for ketone substrate 9 was determined
to be 2.8ꢄ10À6 sÀ1 and the isolated product 11 showed an ee
of 76%. The absolute configuration of amine 11 was assigned
as S by comparison with the (S)-a-methoxy-a-trifluoromethyl-
phenylacetic acid amides of 2-methylpiperidine (4), both
racemic, and the S enantiomer from the reduction of cyclic
imine 2 catalyzed by Streptomyces sp. GF3546 imine reduc-
tase.[20] The imine reductase from S. aurantiacus showed a sig-
nificantly higher catalytic activity than Streptomyces sp. GF3546
imine reductase towards the cyclic imines 1 and 2, however,
led to much lower conversions of ketone substrate 9 in this re-
action setup (Table 2). For other ketone substrates, 12 and 13,
we detected conversions of up to 2.3%, with highest conver-
1
Conversion was determined by H NMR analysis of the crude prod-
uct and ee by chiral-phase HPLC analysis (of the corresponding
acetamides, in the case of 1,2,3,4-tetrahydroisoquinolines 6).
Table 2. Biocatalytic reductive amination of ketones by imine
reductases.[a]
Direct enzymatic amination experiments were performed by incu-
bation of the ketone substrate 9, 12, or 13 (20 mmol) with d-glu-
cose (5 equiv.), NADP+ (10 mol%), GDH (5 U), purified imine reduc-
tase (2.5 mg), and methylammonium buffer (850 mL, 250 mm) in
a total volume of 1 mL. Reactions were terminated after 20 h by
addition of 10wt.% NaOH (100 mL), then extracted with methyl
tert-butyl ether (3ꢄ500 mL) and analyzed by GC–MS.
Ketone
Conversion[b] [%]
Streptomyces sp. GF3546 S. aurantiacus
8.8[c]
1.7[c]
Imine reductase from Streptomyces sp. GF3546 was crystallized by
sitting-drop vapor diffusion from a solution containing 10 mgmLÀ1
of protein at 278 K. Protein solution (0.5 mL) was mixed with the
same volume of a reservoir solution containing 2.0m (NH4)2SO4
and 0.1m BisTris (bis(2-hydroxyethyl)amino-tris(hydroxymethyl)-
methane) buffer at pH 5.5, and equilibrated against the same reser-
voir solution. Needle-shaped crystals were optimized by microseed-
ing, yielding well-diffracting single crystals. The ortholog from
S. aurantiacus was crystallized from 20% (w/v) PEG3350, 0.1m
BisTris buffer at pH 5.5, and 0.25MgCl2·H2O. Diffraction data were
collected at 100 K on beamline X06SA at the Swiss Light Source,
Villigen, Switzerland. The data were integrated and scaled by using
IMOSFLM[21] and SCALA.[22] Molecular replacement was performed
with the model of a putative oxidoreductase from Pseudomonas
0.1[c]
2.3[d]
0.5[e]
1.2[c]
[a] Experiments performed in methylammonium buffer of 212 mm final
concentration containing 20 mmol ketone substrate, 10 mol% NADP+,
and a GDH cofactor regeneration system in a total volume of 1 mL.
[b] Conversion was determined by GC–MS analysis. [c] pH 9.5. [d] pH 9.8.
[e] pH 9.2.
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
[10] a) M. J. Abrahamson, E. Vꢅzquez-Figueroa, N. B. Woodall, J. C. Moore,
A. S. Bommarius, Angew. Chem. 2012, 124, 4036–4040; Angew. Chem.
Int. Ed. 2012, 51, 3969–3972; b) M. J. Abrahamson, J. W. Wong, A. S.
Bommarius, Adv. Synth. Catal. 2013, 355, 1780–1786.
putida KT2440 (PDB 3L6D) and the structures refined by using
REFMAC5.[23]
[11] K. Mitsukura, M. Suzuki, K. Tada, T. Yoshida, T. Nagasawa, Org. Biomol.
Chem. 2010, 8, 4533–4535.
Acknowledgements
[12] K. Mitsukura, M. Suzuki, S. Shinoda, T. Kuramoto, T. Yoshida, T. Nagasa-
wa, Biosci. Biotechnol. Biochem. 2011, 75, 1778–1782.
The authors gratefully thank Hꢁlya Kazanci for technical assis-
tance, Evocatal GmbH for providing GDH, and Dr. Kay Greenfield
for her help in improving the manuscript.
[13] M. Rodrꢆguez-Mata, A. Frank, E. Wells, F. Leipold, N. J. Turner, S. Hart,
J. P. Turkenburg, G. Grogan, ChemBioChem 2013, 14, 1372–1379.
[14] K. Mitsukura, T. Kuramoto, T. Yoshida, N. Kimoto, H. Yamamoto, T. Naga-
sawa, Appl. Microbiol. Biotechnol. 2013, 97, 8079–8086.
[15] S. F. Altschul, W. Gish, W. Miller, E. W. Myers, D. J. Lipman, J. Mol. Biol.
1990, 215, 403–410.
[16] K. Ishida, N. Kimoto, T. Nagasawa, H. Yamamoto, T. Yoshida, Process for
Production of Optically Active Amine Derivative, WO2010024445 A1,
2011.
[17] Shotgun genome sequence of S. aurantiacus JA 4570 DDBJ/EMBL/Gen-
Bank accession number: AOPZ01000000.
[18] F. Sꢅnchez-Sancho, E. Mann, B. Herradꢇn, Synlett 2000, 509–513.
[19] F. Leipold, S. Hussain, D. Ghislieri, N. J. Turner, ChemCatChem 2013, 5,
3505–3508.
Keywords: asymmetric catalysis · biocatalysis · chiral amines ·
dehydrogenase · NADPH-dependent reductase
[1] J. T. Sime, J. Chem. Educ. 1999, 76, 1658–1661.
[2] K. M. Koeller, C.-H. Wong, Nature 2001, 409, 232–240.
[3] T. C. Nugent, M. El-Shazly, Adv. Synth. Catal. 2010, 352, 753–819.
[4] P. Hoyos, V. Pace, A. R. Alcꢅntara, Adv. Synth. Catal. 2012, 354, 2585–
2611.
[5] a) K. R. Bailey, A. J. Ellis, R. Reiss, T. J. Snape, N. J. Turner, Chem. Commun.
2007, 3640–3642; b) V. Kçhler, K. R. Bailey, A. Znabet, J. Raftery, M. Helli-
well, N. J. Turner, Angew. Chem. 2010, 122, 2228–2230; Angew. Chem.
Int. Ed. 2010, 49, 2182–2184; c) D. Ghislieri, A. P. Green, M. Pontini, S. C.
Willies, I. Rowles, A. Frank, G. Grogan, N. J. Turner, J. Am. Chem. Soc.
2013, 135, 10863–10869.
[20] The configuration was confirmed by comparison with authentic (S)-11,
which was obtained via methylation of commercial (S)-1-methyl-3-phe-
nylpropylamine, as suggested by a referee.
[21] H. R. Powell, O. Johnson, A. G. W. Leslie, Acta Crystallogr. Sect. D 2013,
69, 1195–1203.
[6] J.-S. Shin, B.-G. Kim, Biosci. Biotechnol. Biochem. 2001, 65, 1782–1788.
[7] F. G. Mutti, C. S. Fuchs, D. Pressnitz, N. G. Turrini, J. H. Sattler, A. Lerch-
ner, A. Skerra, W. Kroutil, Eur. J. Org. Chem. 2012, 1003–1007.
[8] C. K. Savile, J. M. Janey, E. C. Mundorff, J. C. Moore, S. Tam, W. R. Jarvis,
J. C. Colbeck, A. Krebber, F. J. Fleitz, J. Brands, et al., Science 2010, 329,
305–309.
[22] P. Evans, Acta Crystallogr. Sect. D 2006, 62, 72–82.
[23] G. N. Murshudov, P. Skubꢅk, A. A. Lebedev, N. S. Pannu, R. A. Steiner,
R. A. Nicholls, M. D. Winn, F. Long, A. A. Vagin, Acta Crystallogr. Sect. D
2011, 67, 355–367.
[9] H. Muramatsu, H. Mihara, R. Katutani, M. Yasuda, M. Ueda, T. Kurihara,
N. Esaki, Tetrahedron: Asymmetry 2004, 15, 2841–2843.
Received: April 8, 2014
Published online on && &&, 0000
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
5
&
ÞÞ
These are not the final page numbers!

COMMUNICATIONS
T. Huber, L. Schneider, A. Prꢀg,
S. Gerhardt, O. Einsle, M. Mꢁller*
My oh Streptomyces! S-Selective imine
reductases from Streptomyces are capa-
ble of reducing cyclic imine substrates
and also act as an enzyme catalyst to-
wards imine intermediates derived from
their equilibrium with a ketone and
methylamine. Structural details and
catalytic properties differ among iso-
enzymes of high-sequence homology.
&& – &&
Direct Reductive Amination of
Ketones: Structure and Activity of S-
Selective Imine Reductases from
Streptomyces
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
6
&
ÞÞ
These are not the final page numbers!